Europe's lost forests: a pollen-based synthesis for the last 11,000 years.

8000 years ago, prior to Neolithic agriculture, Europe was mostly a wooded continent. Since then, its forest cover has been progressively fragmented, so that today it covers less than half of Europe's land area, in many cases having been cleared to make way for fields and pasture-land. Establishing the origin of Europe's current, more open land-cover mosaic requires a long-term perspective, for which pollen analysis offers a key tool. In this study we utilise and compare three numerical approaches to transforming pollen data into past forest cover, drawing on >1000 14C-dated site records. All reconstructions highlight the different histories of the mixed temperate and the northern boreal forests, with the former declining progressively since ~6000 years ago, linked to forest clearance for agriculture in later prehistory (especially in northwest Europe) and early historic times (e.g. in north central Europe). In contrast, extensive human impact on the needle-leaf forests of northern Europe only becomes detectable in the last two millennia and has left a larger area of forest in place. Forest loss has been a dominant feature of Europe's landscape ecology in the second half of the current interglacial, with consequences for carbon cycling, ecosystem functioning and biodiversity.

Intercostal muscle action inferred from finite-element analysis.

The external and internal intercostal muscles are important respiratory muscles in humans, but their mechanical actions have been controversial. We used finite-element analysis based on anatomic and mechanical measurements in dogs to assess the action of the intercostal and other rib cage muscles in a model of an isolated canine rib cage. When intercostal muscle forces of either the internal or the external layer were applied in a single interspace, they pulled the adjacent ribs together, consistent with published observations in dogs. However, when the forces were applied in all interspaces, the external layer caused an inspiratory motion and the internal layer caused an expiratory motion, consistent with conventional understanding of intercostal muscle actions. Parasternal intercostal, levator costae, and transversus thoracis (triangularis sterni) muscle actions were also simulated. These muscles caused expected movements of the ribs and sternum. We conclude that the actions of intercostal muscles depend on the spatial extent of their activation. Their actions in a single interspace and in multiple interspaces can be observed and explained with three-dimensional finite-element models.